15:01:22 <adamw> #startmeeting Fedora QA Meeting
15:01:22 <zodbot> Meeting started Mon Mar 11 15:01:22 2019 UTC.
15:01:22 <zodbot> This meeting is logged and archived in a public location.
15:01:22 <zodbot> The chair is adamw. Information about MeetBot at http://wiki.debian.org/MeetBot.
15:01:22 <zodbot> Useful Commands: #action #agreed #halp #info #idea #link #topic.
15:01:22 <zodbot> The meeting name has been set to 'fedora_qa_meeting'
15:01:26 <adamw> #meetingname fedora-qa
15:01:26 <zodbot> The meeting name has been set to 'fedora-qa'
15:01:30 <adamw> #topic Roll call
15:01:35 <tflink> .hello2
15:01:35 <adamw> morning folks, who's around for a qa meeting?
15:01:35 <zodbot> tflink: tflink 'Tim Flink' <tflink@redhat.com>
15:01:40 <frantisekz> .hello2
15:01:40 <lruzicka> .hello2
15:01:41 * kparal good evening
15:01:41 <zodbot> frantisekz: frantisekz 'František Zatloukal' <fzatlouk@redhat.com>
15:01:44 <zodbot> lruzicka: lruzicka 'Lukáš Růžička' <lruzicka@redhat.com>
15:01:50 <lbrabec> .hello2
15:01:51 <zodbot> lbrabec: lbrabec 'Lukas Brabec' <lbrabec@redhat.com>
15:01:58 <coremodule> .hello2
15:01:58 <zodbot> coremodule: coremodule 'Geoffrey Marr' <gmarr@redhat.com>
15:02:04 <coremodule> Hello everyone!
15:02:06 * satellit listening
15:03:56 <Southern_Gentlem> .hello2
15:03:57 <zodbot> Southern_Gentlem: Sorry, but you don't exist
15:04:03 <Southern_Gentlem> .hello jbwillia
15:04:04 <zodbot> Southern_Gentlem: jbwillia 'Ben Williams' <vaioof@gmail.com>
15:04:10 <adamw> man, these ghosts just keep trying to get into the building
15:04:37 <tflink> where's luigi when you need him?
15:04:59 <Southern_Gentlem> or pacman
15:05:16 <adamw> c'mon, guys, who ya gonna call?
15:05:40 <cmurf> i don't believe in no ghost
15:06:10 <adamw> alrighty, thanks for being here, everyone
15:06:11 <tflink> adamw: luigi. definitely luigi - he has plenty of experience capturing ghosts without cities being destroyed in the process :)
15:06:12 <adamw> let's get started...
15:06:18 <adamw> tflink: hah, solid point
15:06:24 <adamw> #topic Previous meeting follow-up
15:06:53 <adamw> #info we have no action items from the previous meeting
15:07:14 <coremodule> lol tflink
15:07:19 <adamw> anyone have any follow-up on anything we did discuss last week?
15:08:48 <adamw> i guess not!
15:09:40 <adamw> #topic Fedora 30 status
15:10:33 <adamw> #info we still had compose issues with Fedora 30 most of last week
15:11:00 <adamw> #info indeed, we still *are* having compose issues: https://pagure.io/dusty/failed-composes/issue/1629 is the latest
15:11:30 <cmurf> is it at least interesting or exciting?
15:12:20 * satellit mate live still has f29 background after  login
15:12:40 <cmurf> oh there are logs
15:13:17 <cmurf> haha is anything not affected?
15:13:24 <adamw> how do you mean?
15:13:34 <adamw> satellit: i'm aware, there's more important stuff atm, like...getting composes
15:13:43 <satellit> +1
15:13:51 <adamw> once we can actually test release-blocking stuff again we can worry about backgrounds in non-releasing-blocking things :)
15:14:33 <Southern_Gentlem> so is the dropping the py2 stuff the cause for the compose issues
15:14:45 <adamw> not at present, no.
15:14:58 <frantisekz> anyway, the dropping should happen in f31 now
15:15:18 <adamw> i think the strongest theory for the most recent failures is 'an evil sorceror hates us and keeps messing with our network connections'.
15:15:27 <adamw> frantisekz: sorry?
15:15:42 <frantisekz> adamw: it was about the py2 dropping
15:15:50 <adamw> oh iswym
15:16:14 <cmurf> i'm gonna guess everything is affected since fedfind isn't finding anything since 20190301
15:16:40 <adamw> #info so far as we know what's in f30 at present works OK, but the lack of composes is a significant impediment to testing
15:16:52 <adamw> not sure there's much else to say on f30 ATM :/ blocker review meeting will be after this meeting
15:16:56 * jskladan_2 lurks
15:16:56 <adamw> any other notes/thoughts on f30?
15:16:58 <cmurf> at least it's failing early in the attempt
15:17:30 <frantisekz> broken spice is quite unfortunate in F30 :/
15:18:45 <adamw> yeah, that one is annoying
15:21:16 <adamw> frantisekz: i'll try and poke some people about it...
15:21:18 <adamw> moving along
15:21:34 <frantisekz> adamw++
15:21:34 <zodbot> frantisekz: Karma for adamwill changed to 11 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
15:21:59 <lruzicka> adamw, without a compose we do not know if it is already fixed or not anyway.
15:22:42 <adamw> lruzicka: sure we do. no-one did anything to fix it, so it's not fixed. :P
15:23:10 <cmurf> haha
15:23:10 <lruzicka> adamw, ok ... I did not know. I am living in KDE at the moment.
15:25:15 <frantisekz> also, adamw, I wanted to discuss a little bit about nomodeset / basic video driver criterion, but we can do that later ( https://bugzilla.redhat.com/show_bug.cgi?id=1683197 )
15:25:24 <adamw> ok
15:25:26 <adamw> #topic Fedora 31 Change review: Rawhide package gating
15:25:35 <adamw> #info https://fedoraproject.org/wiki/Changes/GatingRawhideSinglePackageUpdates
15:25:45 <adamw> so, this is under discussion on devel@ at present
15:26:00 <adamw> i wanted to bring it up here so people are aware of the idea, and to discuss any concerns we have about it from a QA angle
15:27:34 <kparal> looks ok to me
15:29:03 <coremodule> agreed, looks like a good idea
15:29:44 <adamw> just remember, when anything goes wrong with it, the official answer is: #fedora-ci is thataway
15:29:45 <adamw> ;)
15:30:23 <coremodule> haha has anyone seen tflink?
15:30:31 <tflink> ?
15:30:33 <cmurf> ghost got him
15:30:57 <tflink> I don't have any issues with their plans but would re-iterate what adamw said about #fedora-ci
15:31:33 <adamw> alrighty then
15:32:12 <adamw> #agreed we are broadly on-board with the opt-in single package gating plan
15:32:16 <cmurf> guess we'll see if gating turns into damming
15:32:24 <adamw> damning?
15:32:32 <adamw> #topic Release criteria / test case proposal status
15:32:32 <cmurf> nope
15:32:41 <adamw> alright, so, where were we with...stuff
15:32:45 * adamw checks archives hastily
15:33:30 <cmurf> printing proposal sounds straightforward enough
15:34:08 <adamw> #info sgallagh re-proposed the printing criteria last month, and it hasn't received any negative feedback
15:34:34 <adamw> so, one last time or forever hold your peace - anyone concerned about / opposed to the printing criteria as proposed?
15:34:56 <sgallagh> Well, no negative feedback that wasn't a private reply, anyway ;-)
15:35:18 <sgallagh> ... that was a lot of negatives in one sentence.
15:36:42 <adamw> that's okay, we don't have to know about those ;)
15:36:51 <adamw> ...but srsly, anything to worry about?
15:37:03 <cmurf> does QA have an IPP printer?
15:37:15 <cmurf> the two printers are: IPP and print-to-PDF (file)
15:37:38 <sgallagh> cmurf: Does a printer manufactured in the last five years exist that does not have IPP support?
15:37:48 <cmurf> yes
15:37:49 <lruzicka> we do not, I do not think that the office printers are IPP
15:37:50 <cmurf> tons
15:38:27 <sgallagh> lruzicka: The Westford office ones all are, I think.
15:38:36 <sgallagh> I have two in my home office that are.
15:38:41 <adamw> alright, so that seems fine.
15:39:28 <adamw> #agreed there are no objections to the printing criteria as proposed by sgallagh on 2019-02-11, so we will go ahead and implement those
15:39:45 <adamw> #action adamw to work with sgallagh on putting those criteria into practice
15:39:59 <adamw> sgallagh: are there existing test cases for this, do you know? if not, we will need new ones
15:40:13 <sgallagh> I have no idea. I'm far from the expert.
15:40:23 <adamw> i just figured you might've looked
15:40:24 <adamw> alright
15:40:35 <adamw> #action adamw and sgallagh to find or write test cases to back the printing criteria
15:40:46 <lruzicka> lruzicka, they might be, but I do not know, whether they are configured for IPP, Cups says that they are using Foomatic/PostScript
15:40:52 <lruzicka> sgallagh ^
15:41:10 * satellit wireless printers?
15:41:27 <jlanda> too much stack to throubleshooting
15:41:43 <jlanda> I would prefer to avoid wireless printers for this test case
15:41:45 <adamw> i don't think mattdm poked the optical media criterion yet...did you, mattdm?
15:41:56 <adamw> +1 jlanda
15:43:03 <lruzicka> I think that it should be printers that can do PostScript, because otherwise drivers might be troublesome.
15:43:30 <cmurf> lruzicka: that's why the requirement is for an IPP everywhere printer, it's effectively driverless or you could say universal
15:43:52 <sgallagh> Right. The point was to test that the printing stack works.
15:44:02 <jlanda> We can't go more universal than ipp+pdf
15:44:04 <cmurf> that way we don't get into conditional cases like foomatic drivers and such; but yes there are *all* kinds of other ways a particular setup could still fail to print
15:44:12 <lruzicka> cmurf, I do I know from cups.browsed that a printer is IPP configured?
15:44:18 <cmurf> the idea is to make sure the most basic pipeline will work
15:44:23 <sgallagh> The condition is that at least one printer with IPP must print.
15:44:27 <sgallagh> Not that they all must.
15:44:36 <sgallagh> cmurf: Precisely
15:44:36 <lruzicka> because, what my system reports about office printers is Foomatic/Postscript
15:44:46 <cmurf> right although in theory if one works they all should and if not then it's a connectivity issue (or firmware bug maybe)
15:44:51 <sgallagh> We're testing that printing is not *broken for everyone*
15:45:00 <cmurf> exactly
15:45:43 <adamw> #info optical media proposal has yet to be revivified by mattdm
15:45:44 <cmurf> lruzicka: that I'm not sure of off hand, i don't have an IPP printer
15:45:45 <wa1em> linuxmodder lurking
15:45:55 <adamw> any other criteria / test case proposals outstanding that people know of?
15:46:16 <lruzicka> adamw, the modularity test cases that are on test list review ...
15:46:40 <cmurf> sgallagh: huh I wonder if there's a globally available IPP test printer, or if there should be...
15:46:51 <lruzicka> adamw, so far, there were only a couple of objections - kparal suggested to use terminal instead gnome terminal, which I already changed
15:46:51 <sgallagh> lruzicka: I know I owe you a review on those, but I'm badly overcommitted right now
15:46:55 <cmurf> like a virtual printer - i'll investigate
15:47:06 <adamw> ah, right, that's recent
15:47:09 <lruzicka> sgallagh, yes, please, whenever you can ... have a go
15:47:15 <sgallagh> cmurf: If it really comes down to it, I'll make mine available for testing with advance notice.
15:47:23 <adamw> anyone else looked at lruzicka's proposed modularity test cases yet?
15:47:24 <sgallagh> (Set up a 30minute window in my firewall or something)
15:47:40 <lruzicka> adamw, and jlanda suggested not to mention Fedora versions, I deleted them, too
15:47:53 <adamw> that's generally a good idea
15:47:58 <jlanda> I proposed to discard my cron proposal in favour of kparal's approach of server app basic functionalitty, but we need to define a bit to avoid the eternal "what basic functionality is" arguing
15:48:09 <adamw> there are wiki templates you can use for 'current', 'previous', 'next' and 'nextnext' releases, that get updated each release
15:48:20 <adamw> jlanda: oh yeah, that one got sort of stuck there
15:48:38 <lruzicka> adamw, I just wanted to stress that Fedora 28 would be too outdated to test today's status quo of modularity
15:48:50 <adamw> #info lruzicka has modularity test case change proposals out for review on the mailing list at present, please take a look and provide feedback
15:49:02 <jlanda> But anyhow, a big server issue is going to be proposed as blocker, punted to
15:49:10 <jlanda> Fesco and blocked propably
15:49:24 <jlanda> So we could continue with it when we have some more time
15:49:26 <sgallagh> jlanda: Which one is this?
15:49:27 <adamw> #info jlanda's cron proposal produced kparal's "basic server functionality" counter-proposal and things got stuck there, we will wait for one or both of them to move that along
15:49:43 <adamw> sgallagh: i think he means in thory?
15:49:47 <adamw> if one came up
15:50:11 <jlanda> sgallagh: after /etc/cron.d failing prior to fc29 we have been looking for the best way to have a release criteria for those cases
15:50:17 <jlanda> Since we lack one
15:50:17 * sgallagh nods
15:50:41 <jlanda> I proposed an cron specific one, kparal a basic apps approach like in wks
15:50:51 <jlanda> But we are stucked there
15:52:15 <kparal> jlanda: I just presented a few thoughts, it was not really a proposal
15:53:01 <jlanda> we bought your thoughts already, we need to elaborate them :D
15:54:48 <adamw> ok, moving on
15:54:58 <adamw> #topic Test Day / community event status
15:55:09 <adamw> hmm, we seem to be sumantro-less
15:55:14 <frantisekz> :'(
15:55:22 <jlanda> Kernel test is scheduled for tomorrow i think
15:55:48 <frantisekz> yep, it is... are we going to have a test image? I think the kernel in latest F30 compose is older
15:56:01 <adamw> #info there is a kernel test day scheduled for tomorrow: https://fedoraproject.org/wiki/Test_Day:2019-03-12_Kernel_5.0_Test_Day
15:56:17 <cmurf> I see 5.0.1 fc30 in koji
15:56:17 <frantisekz> or we can alter guidelines to update to a kernel from koji
15:56:22 <frantisekz> yeah, but not in repos
15:56:32 <adamw> #info i18n Test Day scheduled for 2019-03-19: https://fedoraproject.org/wiki/Test_Day:2019-03-19_I18N_Test_Day
15:56:46 <adamw> the calendar is showing an IoT test day for Wednesday, but I can't find a wiki page for it
15:56:50 <coremodule> IoT test day is Wednesday https://fedoraproject.org/wiki/Test_Day:Fedora_IoT_Edition_2019-03-13
15:56:51 <pwhalen> https://fedoraproject.org/wiki/Test_Day:Fedora_IoT_Edition_2019-03-13
15:57:01 <adamw> oh, found it
15:57:11 <adamw> #info IoT Test Day scheduled for Wednesday 2019-03-13: https://fedoraproject.org/wiki/Test_Day:Fedora_IoT_Edition_2019-03-13
15:57:18 <adamw> it's missing categories...
15:57:22 <pwhalen> we hope to have a compose to test..
15:57:26 <adamw> yeah
15:57:38 <adamw> if not we may have to pull images out of failed composes, or something
15:57:49 <jlanda> we can change the guideline early tomorrow frantisekz
15:57:52 <adamw> #action adamw to ask sumantro to make contingency plans for upcoming test days if composes are still not available
15:58:15 <adamw> #topic Open floor
15:58:17 <cmurf> speaking of iot there's a website issue, iot.fedoraproject.org goes to the generic apps page
15:58:22 <adamw> so, any other business for the last minute and a half? :)
15:58:30 <frantisekz> so, adamw, do we have a time for that basic video criterion?
15:58:41 <adamw> we've got a minute!
15:58:44 <adamw> what's your concern?
15:58:50 <cmurf> oh yeah, that
15:58:58 <frantisekz> Just wanted to think about it.... as I have a feeling that we are not going to have a fix for "nomodeset" again, same as in F29
15:59:09 <cmurf> well we need a decision if it's supposed to work because as written there's no exception
15:59:12 <frantisekz> if it does make sense to have a criterion and not enforcing it
15:59:17 <cmurf> and it seems to fail on everything that doesn't need it haha
15:59:30 <cmurf> open question if it works on things that do need it
15:59:40 <frantisekz> and maybe, having criterion and possibility to enforce just llvmpipe, but leave modesetting on might make sense
15:59:52 <frantisekz> cmurf: I have no idea how to test that
15:59:59 <cmurf> hardware specific
16:00:14 <jlanda> I haven't be able to found a box that can't properly boot the wks live to a desktop, all my boxes boots properly without nomodeset, and all of them freeze with it :D
16:00:27 <cmurf> all of my hardware has i915 graphics, so modeset 1 works; nomodeset always fails since early Fedora 29 development
16:00:29 <adamw> as written the bug does not actually violate the criterion, i will note
16:00:36 <frantisekz> just for the record, we are talking about https://bugzilla.redhat.com/show_bug.cgi?id=1683197 if anybody does not know
16:00:41 <adamw> the criterion says the menu option has to exist and do what it's supposed to do
16:00:45 <adamw> it doesn't say it has to *work*...
16:00:50 <frantisekz> heh
16:00:54 <cmurf> haha of course
16:00:59 <frantisekz> .... well, fair point .... :D
16:01:04 <adamw> whenever i point this out people laugh, but it's not a joke. that is what it says.
16:01:15 <jlanda> I would be more concerned if we found a box that can't boot with or without nomodeset
16:01:18 <cmurf> best comedy is pointing out the obvious
16:01:27 <cmurf> that everyone else has missed or ignores
16:01:47 <adamw> jlanda: yeah, that's also what i'd be interestedi n
16:01:59 <adamw> pbrobinson told me he had a box that needs nomodeset, i think
16:02:00 <jlanda> I tried, but I didn't found one on my available hardware
16:02:03 <adamw> pbrobinson: any luck testing yet?
16:02:05 <cmurf> i'd prefer a kernel panic if nomodeset is chosen and doesn't work compared to the current behavior
16:02:27 <lruzicka> I would expect that all machines work with nomodeset.
16:02:35 <lruzicka> Seems funny, they do not.
16:02:36 <pbrobinson> adamw: no time as yet, I did get as far as plugging the device in
16:02:44 <cmurf> lruzicka: used to be true, lately not
16:03:00 <lruzicka> cmurf, that is not correct, I believe
16:03:30 <lruzicka> cmurf, that means whenever I have problems with modeset, I am stuck without it as well
16:03:33 <jlanda> Ideally yeah, but in a practical way nomodeset is needed when you have troubles without it
16:04:03 <lruzicka> jlanda, exactly ... but what if it does not work?
16:04:10 <kparal> adamw: "
16:04:10 <kparal> The boot menu for all supported installer and live images should include an entry which causes both installation and the installed system to use a generic, highly compatible video driver (such as 'vesa'). This mechanism should work correctly, launching the installer or desktop and attempting to use the generic driver.
16:04:10 <kparal> "
16:04:20 <kparal> as I read it, it's violating
16:04:22 <jlanda> We're already on blocker meeting time talking here about a blocker bug ;)
16:04:40 <cmurf> kparal \o/
16:04:46 <cmurf> i love that
16:04:49 <adamw> kparal: when it says "this mechanism" it means *literally the boot menu entry itself* must "work correctly"
16:05:04 <adamw> kparal: not that the desktop or whatever must "work correctly"
16:05:10 <kparal> "launching the installer or desktop"
16:05:13 <kparal> which it doesn't
16:05:17 <adamw> sure it does.
16:05:22 <adamw> the desktop doesn't *work*.
16:05:25 <adamw> but it *launches* it. ;)
16:05:26 <kparal> it freezes
16:05:31 <kparal> with black screen
16:05:31 <jlanda> vga=0 works? We could just change the arg :D
16:05:34 <kparal> doesn't launch anything
16:05:36 <adamw> it launches it and it freezes, yes.
16:05:47 <frantisekz> no, it doesn't , at least for me
16:05:47 <cmurf> well it does in fact attempt to use the generic driver (we think but have no confirmation of this) but fails
16:05:54 <frantisekz> stays on plymouth
16:05:55 <adamw> if i launch a rocket and it immediately explodes into a fireball, i still launched it =)
16:05:56 <kparal> I'm surprised how you're trying to sabotage your own criteria
16:06:03 <cmurf> so I agree with adamw it doesn't violate the criteria
16:06:08 <adamw> i'm not, because this is literally what we wanted it to mean, i'm pretty sure
16:06:09 <cmurf> but it's goofy
16:06:19 <adamw> i'd have to go back and dig out the last time we argued about it, though
16:06:21 <cmurf> maybe bring it up on desktop@ there's a thread there about it I think
16:06:26 <jlanda> If it freezes it does not work properly ;)
16:06:38 <kparal> it doesn't make sense to have a criterion that doesn't have the desired effect of allowing people to use safe graphics driver
16:06:46 <jlanda> Er, correctly is the exact word :D
16:06:52 <cmurf> i love it when we have an issue where I agree with everyone
16:06:56 <adamw> anyway, we're way over time at this point
16:07:02 <adamw> so we can continue this next week or on the list
16:07:08 <jlanda> +1
16:07:13 <lruzicka> +1
16:07:30 <cmurf> desktop@
16:07:32 <cmurf> "F30 Testing request: "basic graphics mode" in Workstation"
16:07:35 <cmurf> that's the thread
16:08:17 <kparal> cmurf++
16:08:17 <zodbot> kparal: Karma for chrismurphy changed to 3 (for the current release cycle):  https://badges.fedoraproject.org/tags/cookie/any
16:08:24 <adamw> thanks for coming ,folks
16:08:34 <tflink> adamw: thanks for running the meeting
16:08:42 <kparal> thanks
16:08:46 <adamw> #info some discussion around the usefulness and intent of the basic graphics mode criterion, we will continue on-list or in the next meeting
16:08:49 <adamw> #endmeeting